Microfabricated elastic pillar substrates of various geometries have been widely applied to address many fundamental questions in cell biology regarding the mechanotransduction of cell functions. The tip of pillar is usually fluorescently labeled to obtain high contrast between the pillar and the background of the image. An intensity profile of the fluorescent pillar tips is then modeled by a two-dimensional Gaussian fitting to obtain the pillar position and therefore the pillar deflection produced by cells growing on top of the pillars. Local traction force can be calculated as the produced of the amount of pillar deflection and pillar stiffness. Objective lenses with high magnification (e.g., larger than 60×) are usually used to get a good position resolution of (30˜50 nm of the pillar. However, the coated fluorescent proteins are often non-uniform on a pillar or across pillars. Furthermore, they can be degraded and/or digested by the cells, and dissolved in the media especially during the media swapping. Therefore the quality of fluorescent image is degraded in a time dependent manner, which will affect the accuracy of dynamic cell force measurement. Besides, for objects larger than the optical diffraction limit of the optical system used for imaging, it's not appropriate to use a simple Gaussian fitting to find the center of a pillar since the profile of pixel intensities cannot be described by a Gaussian function. Such an optical system with high magnification objectives have limited field of view and makes it incapable of monitoring large scale concurrent and instantaneous collective cell behavior.